Files
Cerbero-mcp/services/common/src/mcp_common/microstructure.py
T
AdrianoDev a13e3fe045 feat: 15 nuovi indicatori quant (common + deribit + bybit + macro + sentiment)
Common (mcp_common):
- indicators.py: vol_cone, hurst_exponent, half_life_mean_reversion,
  garch11_forecast, autocorrelation, rolling_sharpe, var_cvar
- options.py (nuovo): oi_weighted_skew, smile_asymmetry, atm_vs_wings_vol,
  dealer_gamma_profile, vanna_charm_aggregate
- microstructure.py (nuovo): orderbook_imbalance (ratio + microprice + slope)
- stats.py (nuovo): cointegration_test Engle-Granger + ADF helper

Deribit (+6 tool MCP):
- get_dealer_gamma_profile (net dealer gamma + flip level)
- get_vanna_charm (vanna/charm aggregati pesati OI)
- get_oi_weighted_skew, get_smile_asymmetry, get_atm_vs_wings_vol
- get_orderbook_imbalance

Bybit (+2 tool MCP):
- get_orderbook_imbalance, get_basis_term_structure (futures dated curve)

Macro (+2 tool MCP):
- get_yield_curve_slope (2y10y/5y30y + butterfly + regime)
- get_breakeven_inflation (FRED T5YIE/T10YIE/T5YIFR)

Sentiment (+3 tool MCP):
- get_funding_arb_spread (opportunità arb compatte annualizzate)
- get_liquidation_heatmap (heuristic da OI delta + funding extreme,
  no feed paid Coinglass)
- get_cointegration_pairs (Engle-Granger su coppie crypto Binance hourly)

Tutto in TDD pure-Python (no numpy/scipy in mcp_common). README
aggiornato con elenco completo. 442 test totali verdi.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-27 23:58:07 +02:00

78 lines
2.4 KiB
Python

"""Microstructure indicators: orderbook imbalance, slope, microprice.
Tutte le funzioni accettano bids/asks come list[list[price, qty]] (formato
standard dei ticker exchange) e ritornano metriche aggregate exchange-agnostic.
"""
from __future__ import annotations
def orderbook_imbalance(
bids: list[list[float]],
asks: list[list[float]],
depth: int = 10,
) -> dict[str, float | None]:
"""Imbalance ratio = (bid_vol - ask_vol) / (bid_vol + ask_vol) sui top-`depth`
livelli. Range [-1, +1]. Positivo = bid pressure, negativo = ask pressure.
Microprice (Stoll-Bertsimas): mid pesato dalla size opposta
→ P_micro = (P_bid * Q_ask + P_ask * Q_bid) / (Q_bid + Q_ask). Best level only.
Slope: variazione cumulata di volume per unità di prezzo (proxy per
liquidità in profondità).
"""
if not bids and not asks:
return {
"imbalance_ratio": None,
"bid_volume": 0.0,
"ask_volume": 0.0,
"microprice": None,
"bid_slope": None,
"ask_slope": None,
}
top_bids = bids[:depth]
top_asks = asks[:depth]
bid_vol = sum(q for _, q in top_bids)
ask_vol = sum(q for _, q in top_asks)
total = bid_vol + ask_vol
if total == 0:
ratio = None
else:
ratio = (bid_vol - ask_vol) / total
# Microprice: best bid, best ask. Weighted by opposite-side size.
microprice = None
if top_bids and top_asks:
bp, bq = top_bids[0]
ap, aq = top_asks[0]
denom = bq + aq
if denom > 0:
microprice = (bp * aq + ap * bq) / denom
bid_slope = _depth_slope(top_bids, ascending_price=False)
ask_slope = _depth_slope(top_asks, ascending_price=True)
return {
"imbalance_ratio": ratio,
"bid_volume": bid_vol,
"ask_volume": ask_vol,
"microprice": microprice,
"bid_slope": bid_slope,
"ask_slope": ask_slope,
}
def _depth_slope(levels: list[list[float]], ascending_price: bool) -> float | None:
"""Calcola |Δq / Δp| sul primo vs penultimo livello.
Slope alto = liquidità che crolla rapidamente in profondità (book sottile).
"""
if len(levels) < 2:
return None
p_first, q_first = levels[0]
p_last, q_last = levels[-1]
dp = abs(p_last - p_first)
if dp == 0:
return None
return abs(q_first - q_last) / dp